Getters

DimensionalData.jl defines consistent methods to retrieve information from objects like DimArray, DimStack, Tuples of Dimension, Dimension, and Lookup.

First, we will define an example DimArray.

using DimensionalData
using DimensionalData.Lookups
x, y = X(10:-1:1), Y(100.0:10:200.0)

(↓ X 10:-1:1,
→ Y 100.0:10.0:200.0)

julia> A = rand(x, y)

╭───────────────────────────╮
│ 10×11 DimArray{Float64,2} │
├───────────────────────────┴──────────────────────────────────────────── dims ┐
  ↓ X Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points,
  → Y Sampled{Float64} 100.0:10.0:200.0 ForwardOrdered Regular Points
└──────────────────────────────────────────────────────────────────────────────┘
  ↓ →  100.0       110.0        120.0         …  190.0       200.0
 10      0.616666    0.561953     0.0157247        0.177857    0.856524
  9      0.660024    0.395523     0.0109523        0.552838    0.0536623
  8      0.112606    0.548479     0.0545071        0.755722    0.830655
  7      0.839223    0.0400057    0.702845         0.384116    0.445628
  6      0.734452    0.65906      0.506965    …    0.654591    0.44097
  5      0.89757     0.748725     0.00633515       0.178022    0.106904
  4      0.969026    0.0042232    0.63959          0.423538    0.556096
  3      0.106472    0.274713     0.589536         0.805743    0.390342
  2      0.861535    0.167465     0.507581         0.799964    0.292431
  1      0.245499    0.742732     0.095824    …    0.90203     0.86858

dims retrieves dimensions from any object that has them.

What makes it so useful is that you can filter which dimensions you want, and specify in what order, using any Dimension, Type{Dimension} or Symbol.

julia> dims(A)

(↓ X Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points,
→ Y Sampled{Float64} 100.0:10.0:200.0 ForwardOrdered Regular Points)

julia> dims(A, Y)

Y Sampled{Float64} ForwardOrdered Regular Points
wrapping: 100.0:10.0:200.0

julia> dims(A, Y())

Y Sampled{Float64} ForwardOrdered Regular Points
wrapping: 100.0:10.0:200.0

julia> dims(A, :Y)

Y Sampled{Float64} ForwardOrdered Regular Points
wrapping: 100.0:10.0:200.0

julia> dims(A, (X,))

(↓ X Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points)

julia> dims(A, (Y, X))

(↓ Y Sampled{Float64} 100.0:10.0:200.0 ForwardOrdered Regular Points,
→ X Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points)

julia> dims(A, reverse(dims(A)))

(↓ Y Sampled{Float64} 100.0:10.0:200.0 ForwardOrdered Regular Points,
→ X Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points)

julia> dims(A, isregular)

(↓ X Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points,
→ Y Sampled{Float64} 100.0:10.0:200.0 ForwardOrdered Regular Points)

Predicates

These always return true or false. With multiple dimensions, false means !all and true means all.

dims and all other methods listed above can use predicates to filter the returned dimensions.

julia> issampled(A)

true

julia> issampled(dims(A))

true

julia> issampled(A, Y)

true

julia> issampled(lookup(A, Y))

true

julia> dims(A, issampled)

(↓ X Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points,
→ Y Sampled{Float64} 100.0:10.0:200.0 ForwardOrdered Regular Points)

julia> otherdims(A, issampled)

()

julia> lookup(A, issampled)

Sampled{Int64} 10:-1:1 ReverseOrdered Regular Points,
Sampled{Float64} 100.0:10.0:200.0 ForwardOrdered Regular Points